Mac mini M4 (16GB) personal RAG assistant reachable from your phone
AnythingLLM + Ollama (llama3.1:8b) on a Mac mini M4, exposed through a Cloudflare Tunnel protected by Cloudflare Access, queried from Telegram via n8n.
SoloDocument search / chat with filesUnder $1kSensitivity: Standard
Hardware
| Item | Approx. price (as of 2026-10-16) | Buy |
|---|---|---|
| On sale −14% Apple Mac mini M6 (16GB, 512GB) (Amazon, related current offer) | $949.99 | Check price at Amazon → |
| Apple Mac mini (M4, 16GB RAM, 256GB) | Not stated in source | Check price → |
| Apple Mac mini M4 (16GB, 256GB) (Best Buy, related current price) | $599.00 | Check price at Best Buy → |
Heads up: "Check price" links may be affiliate links. We may earn a commission at no extra cost to you.
Source prices are as stated by the source. Retailer offers: prices checked 2026-10-17, may change. Some links may be affiliate links; we may earn a commission at no extra cost to you. This never changes what we list.
Software stack
- Ollama
- AnythingLLM
- LanceDB
- Cloudflare Tunnel + Access
- n8n (hosted on Hostinger)
- Telegram
Models used
- llama3.1:8b
Good at
- Responses in 3-8 seconds
- 'Costs nothing beyond electricity' (per author)
- Usable from a phone
Weak at
- Queries pass through Telegram, a hosted n8n and Cloudflare - not fully offline
Rough total cost
Hardware price not stated; author reports no running cost beyond electricity (n8n hosting aside).
Getting started
- Install Ollama and pull llama3.1:8b.
- Install AnythingLLM, point it at Ollama, create a workspace and drag in files.
- cloudflared tunnel create rag-tunnel; route your domain to localhost:3001.
- Add a Cloudflare Access service-token policy so only your automation can call it.
- Connect n8n + Telegram bot to the AnythingLLM API.
Source & credit
Building a personal RAG assistant you can actually use — Ovidiu Eftimie (Substack)
Summarized from the source; quotes are the source author's words. Check the original before buying.